This article, originally published Aug. 24, has been updated to include additional information from Complete Genomics.
By Julia Karow
Complete Genomics said this week that it has raised $45 million in private equity from a Series D funding round, which closed six months later than originally planned. As a result, the company, which anticipates reaching profitability next year, has pushed back its plans to launch its human genome sequencing service from June 2009 to January 2010.
The firm said it plans to sequence 10,000 human genomes next year using its proprietary sequencing technology. Previously, it was aiming to sequence 1,000 human genomes by the end of this year and 20,000 in 2010 (see In Sequence 11/4/2008).
The "1,000 genome mark" has "slipped out" by six months, and the company now expects to reach it by mid-2010, Complete Genomics chairman, president, and CEO Cliff Reid told In Sequence this week. The firm is still targeting a price of $5,000 per human genome for batches of genomes when the service launches.
The six-month push-back is tied to the company's inability to raise new funding in a timely manner, he said. In April, Complete Genomics first revealed that the closing of its Series D round was delayed, forcing it to cut costs (see In Sequence 4/10/2009).
"Our timing could not have been worse," Reid said. "We started this financing the day that Lehman Brothers failed and found that in Q4 and Q1, the private equity financing world was really on hold." Only in early April did those companies start to move toward making new investments again, he added.
The new funding, which adds to the $46 million that the company raised in three prior funding rounds, comes from existing investors — Enterprise Partners Venture Capital, OVP Venture Partners, Prospect Venture Partners, and Highland Capital Management — and two new investors, Essex Woodlands Health Ventures and OrbiMed Advisors. Thomas Caskey, an adjunct partner at Essex, and Carl Gordon, a general partner at OrbiMed, joined the firm's board of directors.
The funding will be sufficient to launch the company's human genome sequencing service in January and to build and scale up its commercial sequencing center, based in Mountain View, Calif., over the first six months of 2010 as the firm transitions from an R&D to a services company, according to Reid.
Complete Genomics, which currently has about 120 employees, also plans a “modest headcount increase” prior to the service launch, which Reid said will suffice since “most of our sequencing operations will be automated.” Among the staff, there will be “more software than biochemistry people,” he added.
Its current R&D-type sequencing facility is small, enabling it to conduct pilot projects — typically between five and 10 human genomes at a cost of $20,000 per genome — for early-access customers, of which it has signed up about a dozen, Reid said.
So far, the firm has mentioned the Institute for Systems Biology, the Broad Institute, and the HudsonAlpha Institute as pilot customers, and has said that several pharmaceutical companies are also among them.
For the ISB, Complete Genomics planned to sequence five samples from the same family as part of a larger disease-related study, and for the Broad Institute, a genome "that has already been studied extensively by the scientific community" as well as two pairs of tumor and matched normal samples from a glioblastoma and a melanoma patient.
Over the next few weeks, the company plans to reveal more details about additional customers and their projects, Reid said, and one of its collaborators is scheduled to present a poster at the Personal Genomes conference at Cold Spring Harbor Laboratory next month.
[ pagebreak ]
Scaling Up
The current generation of Complete Genomics' R&D sequencing instruments — a proprietary platform based on short-read sequencing-by-probe-ligation technology — provides about 70 gigabases per run in throughput, and the materials costs per genome are just under $4,000, according to Reid.
The company is able to cover more than 90 percent of a human genome, which is "a good number to enter the commercial market," he said.
Earlier this year, Complete Genomics for the first time publicly showed human genome sequencing data it had generated for a HapMap sample, a project that it plans to publish, according to Reid, though he did not provide an expected timeline for publication (see In Sequence 2/10/2009).
For its January service launch, Complete Genomics plans to install a new generation of sequencing instruments that will have a throughput of about a terabase per paired-end run, each run taking about 12 days. It will also enable the firm to offer its customers large sets of human genomes at a price of $5,000 per genome, a previously stated goal. "We have to get to economies of scale before we can deliver [genomes] at that number," Reid said.
The scale-up of the sequencing center will occur in stages over the first half of 2010, and the total capital cost will be "considerably less" than $50 million, according to Reid, who added that the exact amount is not known yet. "There have been a number of economies of scale that we have been able to exploit to bring the center up for quite a bit less capital than we expected," he said.
The single largest capital purchase will be computing equipment. "We are spending more on computers than we are spending on sequencing instruments and on laboratory automation," Reid said. At present, the company has more than 1,000 compute cores and 1 petabyte of data storage for its own work and pilot projects for customers, and plans to scale this up “significantly” for production, he added. At the time of the service launch in January, it wants to have enough storage and compute power to sequence several hundred genomes per month, he said.
During the first half of 2010, the company expects to be able to produce on the order of 1,000 genomes, he said, but a "very fast ramp-up" will allow it to sequence another 9,000 genomes during the remainder of 2010.
Toward Profitability
Reid said the company expects to become profitable sometime next year. "We are really done with our heavy investments in core technology," he said. "Now we are investing in revenue scale-up and marching toward profitability."
Complete Genomics expects to have customers from two groups: academic researchers, in particular "human genome researchers who are starved for human genome data," and researchers at biopharmaceutical companies trying to understand the genetic basis of diseases or drug responses, according to Reid.
The first "key application" of the service will likely be cancer genome sequencing, he said, "but it's going to be extremely interesting to see what the second and third and fourth applications are going to be."
The company does not plan to offer human genome sequencing services directly to consumers — a service that Illumina started offering earlier this year for $48,000 (see In Sequence 6/16/2009). Rather, it plans to serve as "the back end for any of the direct-to-consumer companies who want to purchase complete human genomes from us and then provide all of the analysis and interpretation and consulting information [for] their consumer customers," Reid said. Those companies would also be responsible for dealing with regulatory and privacy issues arising from working directly with individual customers.
During the next five years, the market will be "primarily driven by the research marketplace," he estimated. After that, "we will see the clinical side overtaking the research side as the most important application ramp for a complete human genome sequence."
[ pagebreak ]
He said he does not anticipate problems arising from privacy concerns over clinical samples. "If you look at the current structure of the industry, you find that most of the major biopharmaceutical firms are already outsourcing their clinical trials to third parties, who essentially have to demonstrate very high competence in maintaining privacy and security," he said. "That's exactly what we will do when the time comes for the clinical applications to become important."
Other researchers, including those involved in the 1000 Genomes Project, have indicated that combining sequence data from several types of platforms yields better results than using data from a single instrument type (see In Sequence 5/12/2009), but Reid does not seem to be concerned about this. “Our objective is to provide such a comprehensive picture of the human genome to our customers that they will not need to use other technologies to gain a complete picture,” he said.
Global Expansion Plans
As previously stated, the company plans to launch several other genome sequencing centers around the world that are modeled after its California center, although it has not decided when. "We have to have this facility run well from an instrument and computing and workflow process of being a highly automated human genome sequencing facility," Reid said. "Once we have that template down, then we will be able to start reproducing it around the world."
The company has seen interest in such sequencing centers from governments of several countries, he said. "We find a number of governments around the world are extremely interested in having a large-scale human genome sequencing center as part of their national technical infrastructure."
Locating such a center in another country would give Complete Genomics access to capital from local partners to cover the building costs, and would enable the company to sequence samples originating in that country without shipping them across borders, Reid said.
He did not name any nations in particular but said that "by having local genome centers in certain countries, particularly within some of the Asian countries, we think the access to DNA will be dramatically enhanced."
Reid added that lower labor costs are not the reason Complete Genomics is planning to open additional genome centers abroad. "Labor cost is such a small component of what we do,” he said. “That's really not the driving force."
According to Reid, Complete Genomics' business model bears some similarity to that of Google. "Google has one type of data — web pages — and it produces one type of output, which is a search report, a sorted list of documents," he said. "We have one type of input — which is human genomes — and we have one type of output, human genome reports, which is a sorted list of variants that describe the differences between the genomes we are sequencing and any other genome. This gives us the opportunity to do a sort of Google-like economies of scale.
"We build this huge back-office infrastructure that is sort of half sequencing and lab automation, and the other half computing, and then enable any researcher in the world with access to Federal Express and the Internet to essentially have their own personal large-scale genome sequencing center available on a moment's notice. Just what you get from the Google search engine we are going to provide to all researchers in terms of sequencing."